Influenza remains a serious cause of illness and death among certain populations. Influenza vaccines in use at present are of limited effectiveness due to the high variability of the virus, and trials all over the world are in progress to enhance their immunogenicity. Conflicting results, in fact, have been reported about the immune response to influenza vaccination in diverse populations. In this paper we analyzed the antibody response to the hemagglutinin (HA) of the H3N2 A/Shangai 16/89 strain, which was included into the trivalent 1991-92 influenza vaccine, in four groups of subjects: 8 healthy young, 13 human immunodeficiency virus (HIV)-infected and 37 elderly healthy people, 9 of whom were treated with thymopentin (TP-5). Our results show levels of anti-HA IgG before vaccination in HIV-infected and elderly people significantly lower than those of normal young subjects. After vaccination, HIV-infected and elderly healthy people showed a significant increase of specific antibodies, whereas a failure in the specific response in normal young subjects was observed, thus differences among the groups were no longer present. Moreover, the spectrotypic analysis of antibody response, by isoelectric focusing and reverse blotting, showed oligoclonal but polymorphic pattern in the majority of subjects, irrespective of the group, and more frequently lack of expansion of the spectrotype after vaccination, thus demonstrating the lack of the recruitment of antigen-specific B cells. Finally, the treatment with TP-5 did not influence the outcome of the vaccination in the group of elderly people. These results further emphasize the limited immunogenicity of influenza vaccination and the inefficacy of TP-5 as immunoadjuvant, in this model of vaccination.